The Spring Hard Drive Guide
- Thread starter pschmid
- Start date
-
- Tags
- Homebuilt NAS / RAID Systems
sandmanwn
Distinguished
- Dec 1, 2006
- 915
- 0
- 18,990
They wont have a 750GB drive until the next revision.
The current platter setup on the 1TB drive is 5x 200GB platters so its impossible to get 750 with the current setup.
The next revision is supposed to be made up of 4x 250GB platters. Which obviously makes it easier to make 500GB and 750GB drives. It should also reduce the manufacturing cost.
I dont remember the estimated arrival of the 250GB platter version but I it shouldnt be long.
The current platter setup on the 1TB drive is 5x 200GB platters so its impossible to get 750 with the current setup.
The next revision is supposed to be made up of 4x 250GB platters. Which obviously makes it easier to make 500GB and 750GB drives. It should also reduce the manufacturing cost.
I dont remember the estimated arrival of the 250GB platter version but I it shouldnt be long.
Alright, I'll be the idiot who tosses RAID into the ring.
2x320GB Seagates yield 640GB at $0.25/GB. Matches the low price in the article, and in RAID0 will outperform anything (except for Raptors, and even then, it should win the sustained read test). And it's >500GB so count it as high capacity.
Of course, if you're up to it -- that is, risking the loss of everything with the failure of just 1 drive -- then I'd reccomend it.
The drive failure argument is valid, but fairly weak. How often has anyone had a single drive system and had the HD fail? Didn't you lose everything? Whatever, statistically, it's 2x the risk of total data loss, but with 2 seperate 320GB partitions I'd have trouble figuring out what to put where...could be why I went with 4x320GB and run a RAID10 dual OS set with a RAID5 data partition (Matrix). Haha, immune from a drive failure and 190MB/s reads on the RAID5 with 130MB/s reads on the RAID10, 120MB/s write on the RAID10 and 60MB/s writes on RAID5.
2x320GB Seagates yield 640GB at $0.25/GB. Matches the low price in the article, and in RAID0 will outperform anything (except for Raptors, and even then, it should win the sustained read test). And it's >500GB so count it as high capacity.
Of course, if you're up to it -- that is, risking the loss of everything with the failure of just 1 drive -- then I'd reccomend it.
The drive failure argument is valid, but fairly weak. How often has anyone had a single drive system and had the HD fail? Didn't you lose everything? Whatever, statistically, it's 2x the risk of total data loss, but with 2 seperate 320GB partitions I'd have trouble figuring out what to put where...could be why I went with 4x320GB and run a RAID10 dual OS set with a RAID5 data partition (Matrix). Haha, immune from a drive failure and 190MB/s reads on the RAID5 with 130MB/s reads on the RAID10, 120MB/s write on the RAID10 and 60MB/s writes on RAID5.
Phrozt
Distinguished
- Jun 19, 2002
- 565
- 0
- 18,980
I like how Pschmid doesn't pay 1 bit of attention to anyone on any of these threads.
I (and others) have asked to see a comparable seagate drive thrown in there because of its PR technology.... but no.. he only includes the biggest and slowest of the seagate drives to represent the name, so the whole thing is distorted as per the current market.
I wish THG would fire this effing idiot.
I (and others) have asked to see a comparable seagate drive thrown in there because of its PR technology.... but no.. he only includes the biggest and slowest of the seagate drives to represent the name, so the whole thing is distorted as per the current market.
I wish THG would fire this effing idiot.
sex_monkey
Distinguished
- May 6, 2006
- 123
- 0
- 18,680
i would like to see how the scores of this lot compare to a WD5000AAKS
western digital 500gb hdd
sorroy to post a link in german but this is the only benchmark i could find. my point being that this hdd is £75 320gb version for £55
http://www.planet3dnow.de/vbulletin/showthread.php?t=301828
western digital 500gb hdd
sorroy to post a link in german but this is the only benchmark i could find. my point being that this hdd is £75 320gb version for £55
http://www.planet3dnow.de/vbulletin/showthread.php?t=301828
disposablehero
Distinguished
- Mar 15, 2007
- 110
- 0
- 18,680
swalkenshaw
Distinguished
- Dec 13, 2006
- 10
- 0
- 18,510
Seagate must not have wanted to pay extra to have more than one drive listed. Did anyone else notice that all the other manufactures have more than one drive listed. THG also only used the ES version of the Seagate drive.
Some may think I love Seagate. I have used many Seagate drives, but they have been having ~10% failure rate at least in the 250 GB 7200.10 series.
Let's see THG do some REAL comparison and use non enterprise versions, and comparable styles. Maybe test them on an AMD server from time to time as well.
Just my 2 cents
Some may think I love Seagate. I have used many Seagate drives, but they have been having ~10% failure rate at least in the 250 GB 7200.10 series.
Let's see THG do some REAL comparison and use non enterprise versions, and comparable styles. Maybe test them on an AMD server from time to time as well.
Just my 2 cents
RambusTech
Distinguished
- Feb 23, 2006
- 22
- 0
- 18,510
nicolasb
Distinguished
- Sep 5, 2006
- 93
- 0
- 18,630
Absolutely!
For some people (HTPC users, for example) lack of noise and vibration is the single most important property of a hard drive, and yet THG doesn't even measure it.
Otherwise this is not a bad review; like previous posters I would like to have seen a few more models reviewed, but that may well be the fault of the manufacturers for not submitting them. Some RAID benchmarks might also have been interesting.
RambusTech
Distinguished
- Feb 23, 2006
- 22
- 0
- 18,510
I have a 15k.5 Cheetah 76gb drive in my computer that makes your raptors look like they're not even spinning.
That is.. if you could tell that they aren't spinning just by looking.. without the casings removed.. but whatever.
That is.. if you could tell that they aren't spinning just by looking.. without the casings removed.. but whatever.
I don't use raptors, and I'm sure your 15,000 rpm drive is very fast.
But like it or not, you are in a rapidly dwindling minority of scsi users
and other than Storage Review I wouldn't expect to see scsi reviews
much anywhere else. That is why I think that, although I agree with some
of the other criticisms leveled here, I can't agree with you because just
not that many people would be interested.
But like it or not, you are in a rapidly dwindling minority of scsi users
and other than Storage Review I wouldn't expect to see scsi reviews
much anywhere else. That is why I think that, although I agree with some
of the other criticisms leveled here, I can't agree with you because just
not that many people would be interested.
Rush job article was it? Nearly missed the deadline?
In page 1 you discuss 2.5" drives then completely forget about them and concentrate on 3.5.
Your look at performance, heat and cost per Gb. But what about power consumption? Noise levels? Warranty?
In page 1 you discuss 2.5" drives then completely forget about them and concentrate on 3.5.
Your look at performance, heat and cost per Gb. But what about power consumption? Noise levels? Warranty?
rockowastaken
Distinguished
- Apr 5, 2007
- 18
- 0
- 18,510
Huh? I guess I'm in too low of an income bracket to be reading THG. Up until I could afford an upgrade a couple of weeks ago, I was deleting data constantly to make room for other data.
My next hard drive gonna be a bunch of flash drives in raid 0 mode.
There are today 30 Gb SATA hard drives emulated with flash memories. All those flash chips running in paralell gonna be the faster hard drive in the west. Is a good bet to believe that they are the hard disk future. Just see the evolution in storage capacity and falling prices. Magnetic hard drives are reaching the end of the histoy.
I officially ranting about lack of this clase of information.
There are today 30 Gb SATA hard drives emulated with flash memories. All those flash chips running in paralell gonna be the faster hard drive in the west. Is a good bet to believe that they are the hard disk future. Just see the evolution in storage capacity and falling prices. Magnetic hard drives are reaching the end of the histoy.
I officially ranting about lack of this clase of information.
"I have a 15k.5 Cheetah 76gb drive in my computer that makes your raptors look like they're not even spinning..."
While I am not disagreeing with that statement, I have a bit of trouble understanding the arithmetic. I do see manufacturers pricing a 15k rpm SCSI way above a 7.2k rpm SATA implying a performance-benefit for the 15k rpm SCSI. A 300gb SCSI drive costs $600 –about as much as 5 (or 6) 500gb ATA drives. Is that real performance, or perhaps the manufacturers trying to milk us if we want the "professional grade" material.
Roughly speaking, a 150gb drive delivers twice what a 75gb drive delivers per spin., but a 15k rpm spins twice for the time a 7.2k rpm drive spin once. From a latency standpoint, the 150gb 7.2k rpm should have about the same latency as a 75gb at 15k rpm. So a drive with over 150gb at 7.2k should have a smaller latency then the 75gb at 15k.
Yes, the SCSI drives typically have better seek time, 8ms range vs the 10-12ms range. That gives another performance boost for the SCSI. Command queuing erased/eased the SCSI command queuing advantage, and SATA300 erased/eased the UltraSCSI 320 bandwidth advantage (and neither can really use all that bandwidth). With 300gb, 500gb, and 750gb ATA at 7.2k, it gets harder for SCSI to beat.
I would imagine a 500 SATA 7.2k runs ring around the 76gb SCSI at 15krpm.
What other factors am I missing? In real-world performance, would you really see a difference? Any real-world benchmarks?
While I am not disagreeing with that statement, I have a bit of trouble understanding the arithmetic. I do see manufacturers pricing a 15k rpm SCSI way above a 7.2k rpm SATA implying a performance-benefit for the 15k rpm SCSI. A 300gb SCSI drive costs $600 –about as much as 5 (or 6) 500gb ATA drives. Is that real performance, or perhaps the manufacturers trying to milk us if we want the "professional grade" material.
Roughly speaking, a 150gb drive delivers twice what a 75gb drive delivers per spin., but a 15k rpm spins twice for the time a 7.2k rpm drive spin once. From a latency standpoint, the 150gb 7.2k rpm should have about the same latency as a 75gb at 15k rpm. So a drive with over 150gb at 7.2k should have a smaller latency then the 75gb at 15k.
Yes, the SCSI drives typically have better seek time, 8ms range vs the 10-12ms range. That gives another performance boost for the SCSI. Command queuing erased/eased the SCSI command queuing advantage, and SATA300 erased/eased the UltraSCSI 320 bandwidth advantage (and neither can really use all that bandwidth). With 300gb, 500gb, and 750gb ATA at 7.2k, it gets harder for SCSI to beat.
I would imagine a 500 SATA 7.2k runs ring around the 76gb SCSI at 15krpm.
What other factors am I missing? In real-world performance, would you really see a difference? Any real-world benchmarks?
I think they have the right idea of how to test the systems and components. maybe because most of their readers are not in the market for the latest & the most expensive immediately it is on the market.
Once they test anything they follow it up 3 & 6 months later to see how it is holding up under normal use. That would be the real life case for most of the end users.
Getting real value for the money by way of decent life expectancy is what testing should be all about.
Once they test anything they follow it up 3 & 6 months later to see how it is holding up under normal use. That would be the real life case for most of the end users.
Getting real value for the money by way of decent life expectancy is what testing should be all about.
TRENDING THREADS
-
-
-
News Firefox user loses 7,470 opened tabs saved over two years after they can’t restore browsing session- Started by Admin
- Replies: 58
Facebook
Twitter
Instagram